Brit. J. vener. Dis. (1959), 35, 71.

ANKYLOSING SPONDYLITIS*BY

R. M. MASONDepartment of Physical Medicine and Regional Rheumatism Centre, The London Hospital

It would seem appropriate to begin this paperby reflecting for a moment on the lady to whom,perhaps, venereologists owe everything, for is it notat least possible that had it not been for Venusherself the need for venereologists would hardlyhave existed? I am indebted to Dr. A. C. Youngfor his historical researches, for he reminds me thatAnchises was a king who spent a night, doubtlessmany nights, with Venus. Unfortunately Zeus foundout about it and threw a thunderbolt at him. Venus,however, so the story goes, was able to interpose hermagic girdle; the thunderbolt was deflected ontohis feet and Anchises, though alive, apparently neverwalked upright again. Venus was married to Vulcan;the Figure shews his kyphotic posture, the fact that

* Paper read to the M.S.S.V.D. on November 28, 1958.

he has to be seated, and that his upper limbs areevidently spared but that he is careful to keep hisheel off the ground.

After classical times a few years seem to havepassed before much interest was shown in theclinical features of spondylitis. It is generally believedthat Marie (1898) and his pupil Leri (1899), Strum-pell (1897), and von Bechterew (1893) provided thefirst adequate descriptions of this disease, butO'Connell (1956) points out that many other descrip-tions preceded those published by these late 19thcentury writers who were unjustifiably rewarded withthe eponymous nomenclature still occasionally usedto-day. Bernard Connor, for instance, in 1695,described a skeleton found in a graveyard and notedfusion of the costovertebral joints and a dorsalkyphosis (Connor, 1695). Lyons (1831) described ak

'Es:.$1

L~~~~~~~~~~~~~~

FIGURE.-Vulcan the Smith at work. Note the kyphosis, that he has to remain seated and that he is careful to keep his heel off the ground.(Courtesy of Dr. A. C. Young.)

71

on August 22, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.35.2.71 on 1 June 1959. D

ownloaded from

2BRITISH JOURNAL OF VENEREAL DISEASES

man of 36 with arthritis of the spine leading tocomplete rigidity in 15 months. According toO'Connell, Lyons tried to arrange the admission ofthis patient to hospital in Dublin with a view toeventual autopsy, pointing out that Hunter had paid£100 for a skeleton with no history whatever whilsthere was one fully documented which, with a littlearrangement, could be had for nothing. The patient,a Manxman, evidently got wind of this and tookhimself off to the Isle of Man to die, leaving strictinstructions to his relatives that after death his bodywas immediately to be cut up into small pieces. Bethat as it may, a year later his bones found their wayback to the Museum of the Royal College ofSurgeons of Ireland, and they are described in themuseum catalogue of that time. Brodie (1850), whofirst described what is now known as Reiter's disease,reported a patient seen in 1841 with pain and stiffnessof the lumbar and dorsal spine, buttocks, and thighs,and arthritis of one knee, who also developed iritis.But to skim a few more years, we find that increas-

ing recognition of the clinical picture, especially inWorld War II and the post-war years, when mostyoung men came under medical supervision as aresult of National Service, has led to a clear charac-terization of ankylosing spondylitis. Hart, Robinson,Allchin, and Maclagan (1949), Hart (1954, 1955),and Baird (1955) in particular have presented uswith a clear-cut clinical picture of an arthritis,always involving the sacro-iliac joints, often thespine, and less commonly the peripheral joints,affecting young men, with an average age at onset of27 years, most commonly in the 15-24 decade andless frequently in the 25-34 and 35-44 decades, nearly10 per cent. having a pre-pubertal onset (Hart, 1955;Mason, Murray, Oates, and Young, 1958).The initial symptom in three-quarters of cases is

low backache or stiffness, stiffness of the groins,thighs, or buttocks, worse in the early morning,often disturbing sleep, and relieved by moving about.Less commonly, the presenting symptom is in thedorsal region, chest, neck, or shoulders, and this isoften when the disease has progressed silently forsome years. For example, a golf professional wasrecently referred with "fibrositis" of the neck. Exam-ination revealed a rigid lumbar and dorsal spine withsome limitation of cervical movements. His anky-losing spondylitis had only revealed itself when thecervical spine had become affected and he was nolonger able to keep his head down while playing golf.In about 10 per cent. of cases the onset is in theperiphery, almost always in the lower limbs, with anarthritis of the ankles, heels, or feet.Ankylosing spondylitis should always be con-

sidered in the differential diagnosis of an inflam-

matory arthritis of the knees, and x rays of thesacro-iliac joints may reveal sacro-iliitis. Clinicalexamination of the patient may show limited move-ments of the lumbar spine, often first affecting lateralflexion. Important additional physical signs includereduced respiratory excursion due to involvement ofthe costo-vertebral joints and tenderness of bonyprominences, especially of the ischium, greater tro-chanters, or symphysis pubis. Forced movements ofthe sacro-iliac joints producing pain is, in ourexperience and that of others (Newton, 1957), amost reliable physical sign. The incidence of iritis,or more precisely unilateral non-granulomatousanterior uveitis, is variously estimated. Hart (1955)found an overall incidence of 13-5 per cent. in hisseries rising to 40 per cent. in those with a historyof 15 years or more and aged over 45 years. Wilkin-son and Bywaters (1958) had an overall incidence of25 per cent., while 36 per cent. of patients withperipheral joint symptoms also had iritis. This com-pares with a much lower incidence of iritis associatedwith rheumatoid arthritis (3 5 per cent.: Duthie,1955) and indeed it is probably true that there is nosignificant association between iritis and rheumatoidarthritis (Stanworth and Sharp, 1956). It seems likelythat the suggestion of an association with rheumatoidarthritis is due either to misdiagnosis or to the con-cept that rheumatoid arthritis and ankylosing spon-dylitis are one and the same disease. It is curiousthat conjunctivitis, not iritis, is associated withReiter's syndrome, at least in the early accounts ofthis disease. However, Paronen (1948) had a 6-6 percent. incidence in his Finnish series, Zewi (1947) 10per cent., and Ford (1953) a 30 per cent. incidence,although (as Stanworth and Sharp point out) Ford'sseries was selected inasmuch as all his patients weresuffering from arthritis of 5 years' or more duration.Iritis may be the presenting symptom, or at least thesymptom which first draws attention to the disease(Teschendorf, 1933; Chan and Sun, 1951; Birkbeck,Buckler, Mason, and Tegner, 1951; Romanus, 1953).Peptic ulceration also appears to be more commonthan in the normal population-some 10 per cent. ofpatients developing clinical or radiological evidence.This complication has a very serious effect ontherapy. That ankylosing spondylitis may also befamilial is now well recognized. Hersh, Stecher,Solomon, Wolpaw, and Hauser (1950) studied theincidence in fifty families and showed that inheritanceis due to an autosomal dominant factor with 70 percent. penetrance in males and 10 per cent. in females;Graham and Uchida (1957) added that when theaffected female is the mother penetrance may beincreased to nearly 100 per cent. One family, whichI have previously briefly reported (Mason, 1951),

72

on August 22, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.35.2.71 on 1 June 1959. D

ownloaded from

ANKYLOSING SPONDYLITIS

illustrates the kind of familial history one mayencounter. The daughter, who presented first, hadtypical ankylosing spondylitis; her brother had back-ache, and on investigation was also found to havethe disease. Both these children admitted that thefather had had backache but, being a fairly toughcharacter, he had never really complained about it.X-ray examination of his spine showed advancedchanges of ankylosing spondylitis.

Recently, a man attended the out-patient depart-ment at the London Hospital with unilateral sacro-iliitis, the nature of which was obscure. We thendiscovered that his brother had also attended thehospital some 2 years previously, and examinationof his x rays showed that he had bilateral sacro-iliitis. It was clear, therefore, that the brother witha unilateral sacro-iliitis was almost certainly sufferingfrom ankylosing spondylitis.

Finally, x-ray evidence of bilateral sacro-iliitis ismandatory to confirm the diagnosis. I do not proposeto discuss in detail the radiological signs, since theyhave been well described not only by Grainger (1957)but by Golding (1936), Hart (1955), and Forestierand Deslous-Paoli (1957).

Characterization of the disease on the basis of thispicture removed from medicine one of the mostcommon causes of diagnostic error, and the oldsaying that every spondylitic had a story to tell, astory of multiple diagnosis, of multiple palliativetherapy, psychiatry, osteopathy, and so on, is nolonger true; indeed, in the R.A.M.C. series, overhalf the cases were diagnosed within 2 years of thefirst symptom (Baird, 1955). If we add to this thewidely-held belief that radiotherapy is effective, wecan sum up the whole story in this equation:

Bilateral sacro-iliitis=Ankylosing spondylitisAnkylosing spondylitis= Radiotherapy

Radiotherapy=Cure.

Such a very simple view deserves the most carefulexamination.

Relationship to Other Diseases

Rheumatoid Arthritis. - Full agreement on thissubject has not yet been achieved. Short, Bauer, andReynolds (1957), from the Massachusetts GeneralHospital in Boston, include ankylosing spondyliticsin their series of patients with rheumatoid arthritis.Collins (1949) says that he can see no pathologicaldistinction between the two diseases, although VanSwaay (1950, 1951) and Cruickshank (1951) distin-guish them on pathological grounds. The cartilagin-ous proliferation which can be seen in ankylosing

spondylitis does not occur in rheumatoid arthritis,there is a tendency to haemorrhage into affectedtissues, and the marked endarteritis obliterans withfrequent bony ankylosis are all distinguishingfeatures. Perhaps the regularly negative serology ofankylosing spondylitis, found not only by ourselves(Mason and others, 1958) but by many other workers(Svartz and Schlossmann, 1950; Ball, 1952; Brown,Badin, and McEwen, 1954; Wilkinson and Bywaters,1958), is the most important and decisive distinguish-ing feature. But sacro-iliitis can occur in rheumatoidarthritis. We have observed unequivocal bilateralsacro-iliitis in some 10 per cent. of our cases ofrheumatoid arthritis, the films being read blind(Mason and others, 1958). Moreover, the peripheraljoint changes of ankylosing spondylitis may be verysimilar both clinically and radiologically, e.g. theperiostitis, the heel erosions, and the simple plantarspurs. These are all non-specific changes, althoughplantar periostitis does not seem to occur in rheuma-toid arthritis and may perhaps be specific for Reiter'sdisease. The peripheral joint changes of rheumatoidarthritis may give a radiological picture indistinguish-able from that of Reiter's disease. Spondylitis canalso occur in rheumatoid arthritis. Sharp (1957) andSharp, Purser, and Lawrence (1958) have very clearlydescribed the differences. In rheumatoid arthritiswith spinal involvement, it is the cervical spine alonewhich is involved, but the cervical spine is never theonly site of ankylosing spondylitis. Complete rigidityof the spine is very rare in rheumatoid arthritis andthe radiological appearances are quite different-there is multiple disk narrowing, it involves the disksabove C.4, there is little or no osteophytosis, thereare erosions of the vertebral plates, vertebral sub-luxation of more than mild degree occurs, and thereis involvement of the atlanto-axial joint. Erosion ofthe apophyseal joints also occurs. In juvenile rheu-matoid arthritis, involvement of the cervical spine iscommon and is characterized by atrophy or hypo-plasia of the vertebrae with ankylosis of the apophy-seal joints.

Rheumatic Fever. - This can also, though rarely,cause a chronic arthritis (Jaccoud, 1869; Poyntonand Pain, 1913; Bywaters, 1950). Thomas (1955)characterized this syndrome as one of recurrentpolyarthritic episodes, rheumatic valvular disease ofthe heart, and atypical spondylitis with peripheraljoint changes. Sharp (1957) also described nine cases.The condition can be distinguished from ankylosingspondylitis by the ligamentous laxity associated withabnormal mobility in the peripheral joints and, inthe spine, by the lack of severe deformity and by theusually normal chest expansion. The sacro-iliac joints

73

on August 22, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.35.2.71 on 1 June 1959. D

ownloaded from

BRITISH JOURNAL OF VENEREAL DISEASES

are abnormal in about half the patients but thechanges are not distinct from those of ankylosingspondylitis, although in the spine itself the antero-posterior narrowing of the vertebral bodies and thenarrowing of multiple disks, with osteophytosis,hypoplasia of the vertebrae, and ankylosis of theapophyseal joints, are characteristic of rheumaticfever.

Psoriasis. Many workers have commented on

the association between psoriasis and ankylosingspondylitis (Sherman, 1952; Sterne and Schneider,1953; Fletcher and Rose, 1955). Wright (1956) foundthree out of 34 cases of psoriatic arthritis to haveankylosing spondylitis, and Sharp (1957) also founda painful limitation of spinal movements in three-quarters of his patients, although he remarked on

the fact that the sacro-iliac joints were often normalin patients with clinical spinal disease in this con-dition. The nails of the psoriatic may, of course, lookrather similar to those seen in cases of keratoderma.

Ulcerative Colitis. Steinberg and Storey (1957)recently described six cases of ankylosing spondylitisassociated with chronic non-specific inflammatorydisease of the intestines. Bywaters and Ansell (1958)found sacro-iliitis in six of 37 cases of ulcerativecolitis with arthritis, the radiological appearancesbeing indistinguishable from ankylosing spondylitis.The association with dysentery is, of course, wellestablished (Paronen, 1948; Reiter, 1916; Fiessingerand Leroy, 1916).

Reiter's Disease.-It is also clear that sacro-iliitisand, indeed, spondylitis (apparently ankylosing) is afairly common sequel to Reiter's disease. Ford (1953)found four cases of ankylosing spondylitis in his 21cases of Reiter's disease and we (Mason and others,1958; Oates, 1958) found that, in over 30 per cent.of our cases of Reiter's disease, there was unequivocalevidence of bilateral sacro-iliitis. In those cases witha history of six years or more, the number rose to54 per cent. Others have commented on this beforeus (Marche, 1950; Sharp, 1957; Thompson, 1957).We have not been able to distinguish any qualitativedifference in the sacro-iliac joints, but severe erosions,marked sclerosis, and particularly ankylosis only seemto occur in Reiter's disease and ankylosing spondy-litis and we have not ourselves seen it in rheumatoidarthritis. However, Sharp (1957) states that he hasobserved fusion of the sacro-iliac joints in rheuma-toid arthritis also. We have mostly been impressedwith the difference in the time of development,sacro-iliitis occurring early in ankylosing spondylitis,and late in Reiter's disease. Sacro-iliitis is also an

occasional feature of brucellosis and Marfan's syn-drome. It is clear, therefore, that the naive belief thatBilateral Sacro-iliitis=Ankylosing Spondylitis doesnot stand up to examination, certainly if we acceptthe concept that there is such a disease entity asankylosing spondylitis.

Treatment

Radiotherapy. - Turning to the second step ofthis simple equation, how does this stand up toexamination? It is a strange thing that this form oftherapy, the possible dangers of which have recentlybeen recognized, should be so readily accepted asbeing effective. If one examines the evidence onwhich this belief is founded, one is startled to findmore evidence of opinion than of data on which tobase such an opinion. Desmarais (1953) carried outone of the few controlled studies of the value ofradiotherapy in this disease. He limited his follow-upto 3 months, but he did find symptomatic relief insome 90 per cent. of the patients given a dose of1000-2000 r. Nevertheless, he noted that 44 per cent.of his controls were also relieved. His numbers weresmall and no treatment group exceeded twenty cases.In addition he found x-ray evidence of progressionof the disease in some of the cases who had obtainedsymptomatic relief. Sharp and Easson (1954) draw aclear distinction between what they call typical andatypical spondylitis in the response to radiotherapy.They found a significant symptomatic response onlyin the typical group, but they too drew attention tothe tendency of the irradiated patients to relapse.Hart (1958), whilst claiming that deep x ray therapycommonly produced clinical improvement, whichmight last for months or years, also said that it wasdebatable whether it altered the natural course of thedisease-"Our own feeling is", he says, "that it doesnot." Wilkinson and Bywaters (1958), in their largeseries, were likewise unimpressed by the real valueof radiotherapy, and found a high relapse rate onadequate follow-up. We must weigh against this therisk of irradiation of the spine-an area containingactive bone marrow to say nothing of the genitalorgans, especially of the female, in the neighbour-hood. Brown and Abbatt (1955) and Abbatt and Lea(1956) have made a very careful statistical study ofthe incidence of leukaemia in men irradiated forankylosing spondylitis, and it seems clear thatleukaemia is significantly more frequent in patientswho have received radiotherapy. In one group thepercentage reached twenty times that to be expected,and it seems that an incidence of about 0 5 per cent.of leukaemia must be expected in patients so treated.Whilst this can hardly be said to be a high incidence

74

on August 22, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.35.2.71 on 1 June 1959. D

ownloaded from

ANKYLOSING SPONDYLITIS

there would appear to be no doubt that by submittingour patients to radiotherapy we are exposing them toa real risk of developing leukaemia-a fatal compli-cation. This calls for at least fairly convincing datathat radiotherapy is more than palliative treatment.

Alternatives to Radiotherapy.-Our own experiencein our Department at the London Hospital has beenthat adequate palliative control of the disease canusually be obtained with phenylbutazone (Buta-zolidin) in small doses of 200 mg. a day, often withas little as 100 mg. a day, combined with mobilizingphysiotherapy, supervision of posture, control ofactivity (mostly encouragement of activity except inthe very active phases), and control of acute exacer-bations, if necessary, by steroids. These measuresreduce the number of cases "needing" radiotherapyto the few resistant cases, perhaps with active pepticulceration, in which surgery is for one reason oranother not practicable, and I believe that even forthis group radiotherapy rarely offers a solution.

Genital Infection.-We have shown (Mason andothers, 1958) that a significantly high proportion ofpatients with typical ankylosing spondylitis haveevidence of chronic prostatitis, and, having carefullyexamined the data, we conclude that this prostatitisis a feature of the disease and not of some otherfactors such as age distribution. Our figures are verysimilar to those of Romanus (1953), who studied 117cases of ankylosing spondylitis but without controls.Nor, indeed, is this a new idea. Running through thewhole literature of ankylosing spondylitis is this re-current theme of genital infection-"gonococcal", ofcourse, before non-specific urethritis was recognized(Marie, 1898; Forestier, 1935, 1939; Volhard, 1938,1943; and many others). Olhagen, Domeij, Giertz,and Romanus (1957) and Romanus (1953) foundthat some 33 per cent. of their patients with anky-losing spondylitis gave a previous history of gonor-rhoea. This work was done in Sweden, however, anddoes not accord with our results in Whitechapel,where we have found a previous history of urethritisin only 9 per cent. of our ankylosing spondylitics-an incidence comparable with that seen in our rheu-matoid group (5 per cent.) whilst 46 per cent. of thepatients with Reiter's disease showed an incidence ofprevious urethritis (Mason and others, 1958).

Attention has already been drawn to the develop-ment of spondylitis in Reiter's disease. How shouldthese findings influence our treatment? Should wesubmit all our spondylitic patients to a determinedattempt to eradicate "infection" from the genito-urinary tract? We should be very careful here. Afterall we have no evidence that genital infection per se

is the cause of typical ankylosing spondylitis. Indeedwe know that in nearly 10 per cent. of these patientsthe onset of the disease occurs before puberty. I amalso impressed with the familial features of thedisease-a familial incidence which does not seemto occur in Reiter's disease. And there are otherdifferences to which I have already alluded-in par-ticular the difference in the time of development ofsacro-iliitis, early in ankylosing spondylitis with asilent prostatitis and late in Reiter's disease with aflorid urethritis.There are at least two fundamental difficulties here:

(1) The complete failure to identify the organismresponsible for Reiter's disease. No doubtthis is a virus for I do not think that theP.P.L.O. can any longer be seriously sus-pected, despite the pioneer work of Dienesand Weinburger (1951) over many years.There is also our complete lack of knowledgeof the aetiology of ankylosing spondylitis.

(2) The second difficulty is perhaps a philosophicone, for, as Bywaters (1958) recently pointedout, we run into difficulty the moment webegin to think of a disease as a real thing, inthe sense, as he says, that a giraffe or ahippopotamus is a real entity. Diseases areno more than reaction patterns to a varietyof stimuli. Now if, with our inadequate,sketchy knowledge, one feels that the reactionpattern in an individual case is characteristicof Reiter's disease then I think that eradi-cation of the genital infection should beattempted, although I should be glad tolearn that this is possible and that there isevidence that by doing so one can influencethe natural course of the disease. If, on theother hand, the reaction pattern is thatwhich we have come to regard as typical ofankylosing spondylitis, then I doubt whetherthe presence of prostatitis is more than oftheoretical interest. However, on this questioncontrolled data are needed. It may well bereasonable to argue on the other hand thatevery effort should be made in fresh cases ofurethritis to eradicate the resulting prostatitisin an endeavour to reduce the number ofcases of spondylitis, and perhaps uveitis, inlater years. What does seem absolutely clearis the tremendous importance of identifyingthe "virus" of Reiter's disease. This wouldindeed represent a major advance in ourknowledge and would put us into the positionwhere some at least of these speculationscould be answered.

75

on August 22, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.35.2.71 on 1 June 1959. D

ownloaded from

BRITISH JOURNAL OF VENEREAL DISEASES

SummaryThe history of the development of our present

concept of ankylosing spondylitis is briefly reviewedand the present day characterization of the diseasedescribed.The various conditions which can give rise to

sacro-iliitis are considered and the place of radio-therapy in the treatment of ankylosing spondylitis iscritically considered.The role of genital infection in ankylosing spondy-

litis, especially in relation to Reiter's disease, isdiscussed.

There are many colleagues whose help I should like toacknowledge, especially that of my Registrars in theDepartment of Physical Medicine at theLondon Hospital.I am also indebted to Dr. A. C. Young for providing theillustration, and to Drs. Stanley Murray and J. K. Oates,who with Dr. Young and myselfformed the team who didmost of the original work discussed here. Finally, I shouldlike to thank Mr. Ambrose King, Senior Physician to theDepartment of Venereal Diseases, the London Hospital,for his constant support and encouragement, and Dr. W.S. Tegner, Physician in Charge, Department of PhysicalMedicine, the London Hospital, for allowing us to usecases under his care.

REFERENCESAbbatt, J. D., and Lea, A. J. (1956). Lancet, 2, 1317.Baird, J. P. (1955). Proc. roy. Soc. Med., 48, 201.Ball, J. (1952). Ann. rheum. Dis., 11, 97.Bechterew, W. von (1893). Neurol Zbl., 12, 426.Birkbeck, H. Q., Buckler, W. St. J., Mason, R. M., and Tegner, W. S.

(1951). Lancet, 2, 802.Brodie, B. C. (1850). "Diseases of the Joints", 5th ed. Longmans,

Brown, Green, and Longmans, London.Brown, W. M. Court, and Abbatt, J. D. (1955). Lancet, 1, 1283.Bywaters, E. G. L. (1950). Brit. Heart J., 12, 101.

(1958). Brit. med. J., 2, 1224.and Ansell, B. M. (1958). Ann. rheum. Dis., 17, 169.

Chan, E., and Sun, S. F. (1951). Chin. med. J., 69, 147.Collins, D. H. (1949). "The Pathology of Articular and Spinal

Diseases." Arnold. London.Connor, B. (1695). Philos. Trans., 19, 21.Cruickshank, B. (1951). Ann. rheum. Dis., 10, 393.Desmarais, M. H. L. (1953). Ibid,. 12, 25.Dienes, L., and Weinberger, H. J. (1951). Bact. Rev., 15, 245.Duthie, J. J. R. (1955). Quoted by Stanworth and Sharp (1956).

Fiessinger, N., and Leroy, E. (1916). Bull. Soc. med. H6p. Paris, 3 s6r.,40, 2030.

Fletcher, E., and Rose, F. C. (1955). Lancet, 1, 695.Ford, D. K. (1953). Ann. rheum. Dis., 12, 177.Forestier, J. (1935). Rev. Rhum., 2, 472.

(1939). Radiology, 33, 389.and Deslous-Paoli, P. (1957). Ann. rheum. Dis., 16, 31.

Golding, F. C. (1936). Brit. J. Surg., 23, 484.Grainger, R. G. (1957). Proc. roy. Soc. Med., 50, 854.Graham, W., and Uchida, I. A. (1957). Ann. rheum. Dis., 16, 334.Hart, F. Dudley (1954). Ibid., 13, 186.

(1955). Ibid., 14, 77.(1958). J. Indian med. Prof., 5, 2166.Robinson, K. C., Allchin, F. M., and Maclagan, N. F. (1949).Quart. J. Med., 18, 217.

Hersh, A. H., Stecher, R. M., Solomon, W. M., Wolpaw, R., andHauser, H. (1950). Amer. J. hum. Genet., 2, 391.

Jaccoud, S. (1869). "LeCons de clinique m6dicale faites a l'Hopital dela Charit6", 2nd ed. Delahaye, Paris.

L-ri, A. (1899). Rev. Med. (Paris), 19, 597.Lyons, P. M. (1831). Lancet, 1931-32, 1, 27.Marche, J. (1950). Rev. Rhum., 17, 449.Marie, P. (1898). Rev. Med. (Paris), 18, 285.Mason, R. M. (1951). Ann. rheum. Dis., 10, 78.

,Murray, R. S., Oates, J. K., and Young, A. C. (1958). Brit. ined.J., 1, 748.

Morrison, R. J. G. (1955). Proc. roy. Soc. Med., 48, 204.Newton, D. R. L. (1957). Ibid., 50, 850.Oates, J. K. (1958). Brit. J. vener. Dis., 34, 177.O'Connell, D. (1956). Ann. rheum. Dis., 15, 119.Olhagen, B., Domeij, G. Giertz, C. and Romanus, R. (1957). "XI

International Rheumatological Conference, Toronto, Canada",p. 63.

Paronen, I. (1948). Acta med. scand., 131, Supp. 212.Poynton, F. J., and Paine, A. (1913). "Researches on Rheumatism".

Churchill, London.Reiter, H. (1916). Dtsch. med. Wschr., 42, 1535.Romanus, R. (1953). Acta med. scand., 145, Suppl. 280.Sharp, J. (1957). Brit. med. J., 1, 975.

and Easson, E. C. (1954). Ibid., 1, 619.Purser, D. W., and Lawrence, J. S. (1958). Ann. rheum. Dis., 17,303.

Sherman, M. S. (1952). J. Bone Jt Surg., 34A, 831.Short, C. L., Bauer, W., and Reynolds, W. E. (1957). "Rheumatoid

Arthritis". Harvard University Press, Cambridge, Mass.Stanworth, A., and Sharp, J. (1956). Ann. rheum. Dis., 15, 140.Steinberg, V. L., and Storey, G. (1957). Brit. med. J., 2, 1157.Sterne, E. H., and Schneider, B. (1953). Ann. intern. Med., 38, 512.Strumpell, A. (1897). Dtsch. Z. Nervenheilk., 11, 338.Svartz, N., and Schlossman, K. (1950). Ann. rheum. Dis., 9, 377.Swaay, H. Van (1950). "Spondylosis Ankylopoetica", p. 159. Edward

Ijdo, Leiden.(1951). "II Congr6s Europ6enne de Rhumatologie", Barcelona.

Teschendorf, H. J. (1933). Dtsch. med. Wschr., 59, 1576.Thomas, A. E. (1955). Ann. rheum. Dis., 14, 259.Thompson, M. (1957). Proc. roy. Soc. Med., 50, 847.Volhard, E. (1938). Z. Rheumaforsch., 1, 481.

(1943). "Entzundliche Wirbelsaulenversteifung (Morbus vonBechteren-Strumpell-Marie)." Barth, Leipzig.

Wilkinson, M., and Bywaters, E. G. L. (1958). Ann. rheum. Dis., 17,209.

Wright, V. (1956). Ibid., 15, 348.Zewi, M. (1947). Acta ophthal. (Kbh.), 25, 47.Ziff, M., Brown, P., Badin, J., and McEwan, C. (1954). Bull. rheum.

Dis., 5, 75.

76

on August 22, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.35.2.71 on 1 June 1959. D

ownloaded from